Google 发布 Gemini 4，称其能力强大，目前仅向“受信任的网络防御者”开放
英文标题：Google announces Gemini 4 and says it’s so capable that only ‘trusted cyber defenders’ can have it right now
Google 发布了新一代前沿 AI 模型 Gemini 4 Argon，称其在真实软件工程、法律与金融等企业知识工作以及网络安全防御的复杂工作流中具备前沿性能。该模型目前仅限一组“受信任的网络防御者”使用，Google 正参与美国政府的模型发布前访问自愿流程，并将逐步扩大开放范围。Google 表示，Gemini 4 Argon 已用于内部工作流，例如大规模代码库迁移，并在多项基准测试中优于 OpenAI 和 Anthropic 的竞品模型。该发布紧随 OpenAI DevDay 之后，反映出前沿模型竞争正进一步延伸至 Agent、软件工程与安全能力领域。
273.3 KB
56.5 KB
3.1 KB
598.9 KB
65.1 KB
112.3 KB
619.2 KB
407.0 KB
472.3 KB
868.5 KB
47.6 KB
95.2 KB
972.7 KB
258.2 KB
439.2 KB
24.2 KB
中文
Google 今天公布了其下一代前沿 AI 模型，命名为 Gemini 4 Argon。根据 Google 首席 AI 架构师兼 Google DeepMind 高级副总裁 Koray Kavukcuoglu 的说法，这款新模型可在真实软件工程、法律和金融等企业知识工作，以及网络安全防御领域的复杂工作流中，提供“前沿级性能”。但该公司初期仅向“一组受信任的网络防御者”开放访问；Kavukcuoglu 表示，Google“正在积极参与美国政府针对模型发布前访问的自愿流程，同时逐步扩大访问范围”。
Kavukcuoglu 表示，Gemini 4 Argon 已经在支持 Google 的“内部工作流”，并特别提到该模型已帮助完成诸如大规模代码库迁移等任务。Google 还发布了一张大型图表，展示 Gemini 4 在多项基准测试中优于 OpenAI 和 Anthropic 的竞争模型。（本周早些时候，AI 社区曾在 X 上热议疑似泄露的新模型基准测试结果。）在更广泛推出之前，Google 将加强“关键前沿安全保障措施”，包括防范滥用和提示注入攻击，以及监测模型失调问题。
Gemini 4 Argon 的发布距离 OpenAI 高调举办的 DevDay 大会仅过去一天；OpenAI 在此次活动中推出了其 Dots AI Agent和 GPT-6.1 Sol AI 模型。（OpenAI 本周还表示，出于对安全问题的担忧，将不会发布原计划推出的 GPT-6.1 Astra 模型。）此次发布也发生在 Kavukcuoglu 于8 月被任命为 DeepMind 新负责人之后不久。
English / 原文
Google today revealed its next AI frontier model, which it’s calling Gemini 4 Argon. The new model delivers “frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense,” according to chief AI architect and Google DeepMind SVP Koray Kavukcuoglu. But the company is limiting access at first to a “set of trusted cyber defenders,” and Kavukcuoglu says that Google is “actively engaged in the U.S. government’s voluntary process for pre-release model access while we gradually expand access.”
Kavukcuoglu says Gemini 4 Argon is already powering Google’s “internal workflows,” highlighting how the model has helped with tasks like large-scale codebase migrations. Google also shared a big chart showing many benchmarks where Gemini 4 fares better than competing models from OpenAI and Anthropic. (The AI community was buzzing earlier this week on X over what appeared to be leaked benchmarks for the new model.) Google will be strengthening “critical frontier safeguards” before a broader rollout, including defending against misuse and prompt injection attacks and monitoring for misalignment.
The launch of Gemini 4 Argon comes just a day after OpenAI’s splashy DevDay conference, where OpenAI launched its Dots AI agent and GPT-6.1 Sol AI model. (OpenAI also said this week that it wouldn’t be releasing a planned GPT-6.1 Astra model due to worries about safety.) It’s also arriving soon after Kavukcuoglu’s appointment as the new boss of DeepMind in August.